The project involves research on the processing of speech and complex sounds at three levels of the auditory system: 1. Peripheral stages - theoretical models of cochlear function to explain the results of various physiological and psychophysical experiments. 2. Neural network models to process and extract important parameters of speech and other sounds for both monaural and binaural hearing. 3. Learning algorithms mimicking adaptive central auditory neural networks to perform storage and recognition tasks.